Re: Dictionary to translate (was:Re: Persian translations)

[Ramanan Selvaratnam <rama uklinux net>]

On Mon, 2003-12-01 at 17:01, Pablo Saratxaga wrote:

> > (Did  I mention the many keyboard layouts?)
> 
> That is not a problem, as it is only a concern for the user,
> intercomunication is independent of keyboard layouts; you don't
> need to know, nor care about, how someone inputs his text;
> the way how they are exchanged however is important.

True. But as you will know best ... at the moment the Tamil development
has been like ... 

[right click]
	Input Methods
		Tamil xyz (Unicode)
		Tamil xyz  (Non standard enc. 1)
		Tamil xyz  (Non standard enc. 2)
		...
		Tamil abc (Unicode)
		Tamil abc  (Non standard enc. 1)
		Tamil abc  (Non standard enc. 2)
		...
		so on...

I never understood how encoding specific keyboard layouts came about
(again no docs) but shouldn't we actively promote 

[right click]
	Input Methods
		Tamil xyz
		Tamil abc
		...

whereby UTF-8 is what gets used.

Actually with a growing list of langs into GTK-IM this will be more like
[right click]
	Input Methods
		Dravidian|| South Indian
			Tamil xyz

In fact Dasher and similar should get in there.

I have been having good results with the Dasher I am training to speak
Tamil. 3.2.0 inputs text into any UTF-8 app.
all it needs is  good training text 
http://www.inference.phy.cam.ac.uk/dasher/Training.html


> > I think an electronic dictinary is the best solution.
> 
> and a spellchecker (for languages for which it is easily done; if 
> Tamil uses blanks as word separators and is not agglutinative, it's ok).

Very interesting. I will mail you seperately on this.

Back to the topic (Michael Meek's mail to the list actually throws more
light on the urgency to better understand the copyrights)

> 
> > In the case of Tamil I am not too far from a very good one maintained by
> > the state university.
> > 
> > How will the copyrights and licensing affect this?

> It must be freely usable, copiable and distributable, clearly.

GPL or compatible. Correct?

Maybe you would care to discuss this on the other thread which to me is
building up to what is translation.

> Now, for freely modifiable, it can be argued (if some official endorsement
> is needed, as is the case in Iran, it won't be fully freely modifiable
> for example); but the abilit to fully use and spread it without any
> restraint is necessary to make it useful.

This is not GPL or compatible. Correct?

> note I talk about a basic english/tamil dictionnary, or even a list
> of tamil words; a full fledged dictionnary, with lenghty definitions,
> may be different... or not; look at http://ta.wikipedia.org/ for a project
> of a free (GFDL licence) encyclopedia in Tamil language.

You know what, first time I am seeing this. It only fortifies my point
that the existing FS infrastrucutre is more than adequate for free
software advocates to continue the developments globally.
Of course reference points where central coordination nearer to the end
users are needed but what is the point if I have been subscribed to
every "Linux" interest and related Tamil developments like fonts and
encoding, being active in a such a way it is hard to miss me and then I
learn of this great development here? 
Many thanks for pointing me here though.
This probably changes plans yet again for me to the good :-)


> it would be nicer if public governemental efforts were done from the
> beginning with the idea of producing free content; it will even means
> less costs for them. 
True. I like the term free information though.

Best regards,

Ramanan